Summary: | 碩士 === 國立交通大學 === 電機與控制工程系 === 88 === Abstract:
This thesis aims to investigate the performance of an image coding method, operational optimal Vertex-Based Shape Coding (VBSC), applied to both the boundary coding and EEG (electroencephalograph) data compression. Boundary represen- tation is one of the tasks often required in the post-pro-
cessing of a contour-type image. The conventional chain-code descriptor provides a simple way to represent a boundary, yet the compression ratio is low. Compression efficiency and per- formance of the VBSC introduced in this thesis highly depend
on the implementation parameters.
In this study, we first investigate the effect of two para- meters, the admissible control point width DM and the allowed distortion Dmax, on the compression efficacy and quality of the decoded boundary. In addition, a quantitative approach for mea-suring the complexity of a boundary is presented. Performance of the VBSC on several boundaries of different degrees of com- plexity is investigated. Finally, the method is applied to the EEG data compression.
|